Zoological Studies 43(1): 112-122 (2004)
Is Central Mountain Range a Geographic Barrier to the Giant Wood
Spider Nephila pilipes (Araneae: Tetragnathidae) in Taiwan? A
Population Genetic Approach
Jun-Wei Lee1, Ling Jiang2, Yong-Chao Su1 and I-Min Tso1,*
1Department
2Department
of Biology, Tunghai University, Taichung, Taiwan 407, R.O.C.
of Biology, National Changhua University of Education, Chunghua, Taiwan 500, R.O.C.
(Accepted October 6, 2003)
Jun-Wei Lee, Ling Jiang , Yong-Chao Su and I-Min Tso (2004) Is the Central Mountain Range a geographic
barrier to the giant wood spider Nephila pilipes (Araneae: Tetragnathidae) in Taiwan? A population genetic
approach. Zoological Studies 43(1): 112-122. Most phylogeographic studies in Taiwan show that the Central
Mountain Range is a major geographic barrier to vertebrates inhabiting low-elevation areas. In this study, we
choose to investigate the widely distributed giant wood spider Nephila pilipes (Fabricius, 1793) to determine if
their population genetic structure also shows an east-west differentiation pattern resembling those of terrestrial
vertebrates studied so far. Mitochondrial cytochrome oxidase I (COI) was used as a genetic marker, and its
partial sequence was determined in 189 specimens collected from 24 localities in the Ryukyu Islands, Taiwan,
and mainland China. The 617-base partial sequence of COI was determined from DNA extracted from the leg
muscle, and 11 haplotypes were identified from all specimens examined. Neighbor-joining (NJ) and maximum
parsimony (MP) methods were used to construct phylogenetic trees using N. clavata Koch, 1877 and N. antipodiana (Walckenaer, 1841) as outgroups. Results from both methods indicate that N. pilipes populations can be
separated into 3 major lineages: group A (haplotypes EA, RK, CN2, CN4, CN5, and CN6), B (haplotypes TW1,
TW3, CN1, and CN3), and C (TW2). Group A consists of most specimens from 23 localities. Group B consists
of specimens from southeastern China and northwestern Taiwan. Group C consists of a few specimens from a
single locality in northeastern Taiwan. The percentage sequence differences from pairwise comparisons of all
haplotypes ranged between 0.2% and 3.5%. Within-region nucleotide diversity (π) ranged between 0.0% and
0.57%. The EA haplotype was the main component of all populations, and haplotypes in different Taiwanese
populations were not structured geographically. Haplotypes TW1, 2, or 3 were sporadically distributed and
could only be found within a few populations. These results indicate that a high level of gene flow exists among
different populations of N. pilipes in Taiwan, and therefore the Central Mountain Range does not seem to be a
major geographic barrier to this spider. http://www.sinica.edu.tw/zool/zoolstud/43.1/112.pdf
Key words: Nephila pilipes, Population genetic structure, Mitochondrial DNA, Cytochrome oxydase I, Taiwan.
H
disperse into Taiwan and to separate from ancestral populations (Shen 1997). The CMR, which
was formed under the effect of tectonic dynamics
and runs along the north-south longitudinal axis of
the island, has been suggested to be a geographic
barrier to many species of terrestrial vertebrates
inhabiting low-elevation areas. This geographic
barrier further facilitates the differentiation of invading populations. For example, on the basis of
,
PCR-RFLP data, Yeh (1997) found that Moltrechti s
treefrog Rhacophorus moltrechti in Taiwan could
istorical vicariance events during glacial
periods and the geographic barrier provided by the
Central Mountain Range (CMR) have long been
considered to be 2 major factors affecting the
genetic structures of several species of vertebrates in Taiwan (Lue 1998). Taiwan was connected with and separated from the Asian continent
several times due to sea-level changes during
glacial and interglacial periods (Lin 1963 1966).
Each connection and isolation provided terrestrial
animals on the Asian mainland an opportunity to
*To whom correspondence and reprint requests should be addressed. Fax: 886-4-23590296. E-mail: [email protected]
112
Lee et al. -- Population Genetics of Giant Wood Spiders
be divided into 2 distinct (eastern and western) lineages. Liu (1995) also found a similar pattern in
,
Swinhoe s tree lizard Japalura swinhonis by restriction fragment length polymorphism. Using
allozyme electrophoresis, Toda et al. (1998) found
substantial differences in allele frequencies at several loci between the eastern and remaining populations of the Indian rice frog Rana limnocharis.
These studies demonstrate that many terrestrial
vertebrates can be divided into eastern and western groups according to their population genetic
structures. The barrier provided by the steep CMR
might be responsible for these shared differentiation patterns.
Although the population genetic structures of
several terrestrial vertebrates in Taiwan are well
understood, those of terrestrial arthropods are still
largely unknown. It is not clear whether or not historical vicariance events and the CMR have affected terrestrial arthropods in the same way as those
events affected land vertebrates. The population
genetic structure of Psechrus spiders (Araneae:
Psechridae) inhabiting low-elevation forests of
Taiwan was examined by Lin et al. (1999). Based
on RAPD data, Lin et al. (1999) reported that
Taiwanese Psechrus populations can be separated
into northern/central and southern/eastern lineages, a pattern similar to those of vertebrates
investigated so far. However, the Psechrus specimens examined in that study exhibited great variation in genital structures. The authors could not be
certain whether the specimens used in the phylogenetic analysis were members of a single species
or several species. Therefore, the conclusions of
Lin et al. (1999) remain controversial, and similar
studies using taxa with confirmed taxonomic status
are still needed.
The aforementioned historical events and
geographic barrier might have affected terrestrial
arthropods, especially those inhabiting low-elevation forests, in ways quite different from vertebrates. Take the giant wood spider Nephila pilipes
(Fabricius, 1793) (Araneae: Tetragnathidae) as an
example. Nephila pilipes is abundant in low-elevation subtropical and tropical forests of Taiwan (Lee
1964). The mean annual temperature of the current northern most distribution range of N. pilipes
in East Asia (in southern Japan; Yaginuma 1986)
is about 18 C (Shen 1996). During glacial periods
of the Pleistocene, however, the average temperature of most regions in Taiwan ranged from 8.0 to
11.0 C (Lin and Lin 1983). Furthermore, most
areas were occupied by either savanna (Tsukada
1966) or pine forests (Shen 1996), which were
°
°
113
unsuitable for N. pilipes. It is quite possible that N.
pilipes became extirpated from Taiwan during the
glacial periods of the Pleistocene. Current populations might have entered Taiwan during the latest
glacial period, which occurred 15 000 to 10 000 yr
ago. Individuals entering Taiwan after the last
glacial period might have rapidly reoccupied and
eventually have become distributed throughout the
island. Such a short period of time may not have
allowed differentiation of populations inhabiting
western and eastern sides of the CMR to have
been built up. Therefore, for N. pilipes or terrestrial arthropods inhabiting low-elevation areas in
general, a more-homogenous population genetic
structure should be expected.
This study examines the genetic structure of
the giant wood spider Nephila pilipes to determine
if populations in Taiwan exhibit an east-west differentiation pattern similar to those of vertebrates, or
a homogenous pattern according to our hypothesis. In this study, in addition to specimens from
Taiwan, we also obtained a limited number of
specimens from the southeastern coast of mainland China. Although phylogeographic studies of
terrestrial animals in Taiwan are abundant, most of
them have only examined Taiwanese materials
(Tzeng 1986, Yang et al. 1994, Tsai 1999, Yang
1999, Hsu 2000 2001, Wang et al. 2000). Only a
very few of them incorporated specimens from
mainland China in the analysis (e.g., Toda et al.
1997). Evidence illustrating how historical vicariance events determine phylogenetic relationships
of terrestrial animals co-inhabiting southeastern
China and Taiwan is still quite limited. Therefore,
another objective of this study was to compare the
haplotypes of Taiwanese and mainland China populations to gain a preliminary assessment of their
phylogeographic relationship.
MATERIALS AND METHODS
Sampling localities
One hundred eighty-nine female specimens
of Nephila pilipes were obtained from 18 localities
in Taiwan and nearby coastal islands, 2 localities
from the Ryukyu Is., and 4 localities from southwestern mainland China (Table 1; Fig. 1). More
than 50 km separated each locality. In every locality, spiders were sampled from an area greater
than 2 km2 to avoid obtaining offspring of the same
breeding individuals. Female N. clavata Koch,
1877 (GenBank accession no. AY052586) collect-
114
Zoological Studies 43(1): 112-122 (2004)
ed from Maolin, Kaohsiung County, Taiwan and
female N. antipodiana (Walckenaer, 1841)
(GenBank accession no. AY052587) collected
from Singapore were both used as outgroups in
the subsequent analyses.
DNA extraction, amplification, and sequencing
Table 1. Collection localities, haplotypes, and
GenBank accession numbers of Nephila pilipes
examined in this study. Numbers in boldface indicate the number of specimens examined in each
population
Code
Locality
1
2
3
Iheya I. Japan
Iriomote I. Japan
Guangdong, China
4
5
6
7
8
9
Sample size Haplotypes Accession no.
Guangxi, China
Yunnan, China
Fujian, China
Shuangshi, Taiwan
Nankang, Taiwan
Keelung, Taiwan
10 Foyanshan, Taiwan
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Huisunlinchang, Taiwan
Lianshan, Taiwan
Dajin, Taiwan
Kenting, Taiwan
Nanjenshan, Taiwan
Shuanglianpi, Taiwan
Gululindao, Taiwan
Kueishan I., Taiwan
Fuyuan, Taiwan
Baqi, Taiwan
Shenmigu, Taiwan
Chipon, Taiwan
Green I., Taiwan
Orchid I., Taiwan
4
2
10
3
3
1
3
1
6
3
2
1
13
10
3
9
6
9
7
2
21
16
5
4
5
8
4
6
23
19
3
1
10
2
6
5
5
9
10
11
RK
RK
AY052595
EA
CN1
CN2
CN5
CN3
AY052594
AY052588
AY052589
AY052592
AY052590
EA
CN1
CN4
EA
CN6
EA
EA
EA
TW1
°
°
AY052591
AY052593
AY052596
EA
TW1
EA
EA
EA
EA
EA
EA
TW2
TW3
EA
EA
EA
EA
EA
EA
EA
EA
Specimens used for DNA extractions were
either frozen at -80 C or were preserved in 95%
ethanol. Crude DNA was extracted from the muscle of legs or pedipalps using the phenol/proteinase K/sodium dodecyl sulfate method
(Sambrook et al. 1989). The mtDNA cytochrome
oxidase I (COI) partial sequence was amplified
using the polymerase chain reaction (PCR), with a
set of universal primers (Folmer et al. 1994) modified on the basis of consensus sequences of several arthropods (Crustacea X69067, Araneae
X74571, and Insect NC001322) retrieved from
NCBI GenBank (Bethesda, Maryland, USA): L CO
,
,
1490 5 -GGTCAACAAATCATAAAGATATTGG
-3 ;
,
H CO 2198
5 - TAAACTTCAGGGTGACCAAAAA,
ATCA -3 ; with L standing for light and H standing
for heavy DNA strands, CO referring to cytochrome
oxidase, and the numbers (1490 and 2198) repre,
senting the position of Drosophila yakuba s mitochondrial genome (Folmer et al. 1994). Bold and
italicized letters indicate modifications from the
original sequences.
Amplification was carried out for 30 cycles as
follows: 70 s of denaturation at 95 C, 90 s of
annealing at 58 C, and 2 min of extension at 72 C,
followed by a final 10-min extension at 72 C.
PCR products were separated and eluted using
agarose gel purification (Gene-Spin TM Gel
Extraction Kit, Protech Technology, Taiwan). Cycle
sequencing with Taq polymerase was carried out
with an ABI PRISM BigDye Terminator Cycle
Sequencing Ready Reaction Kit (PE Biosystems,
U. S. A.), and it was directly sequenced using an
automated DNA sequencer (ABI PRISM TM 377
DNA Sequencer, Perkin Elmer, U. S. A.). For
sequencing, the same PCR primers were used.
The sequence of each sample was verified
through a comparison of light and heavy strands.
°
°
°
Sequence alignment and phylogenetic analyses
AY052597
AY052598
All sequences were aligned using the Clustal
W program (Thompson et al. 1994) in the BioEdit
vers. 4.7.8 (Hall 1999) computer software and corrected manually using the GeneDoc (Nicholas et
al. 1997) computer software. Phylogenetic trees
were constructed by both the distance matrix
method (Neighbor-joining (NJ) method; Saitou and
Nei 1987) and the character state (maximum parsimony; MP) method. The NJ analysis was per,
formed by calculating Kimura s (1980) 2-parameter
distance with the transition/transversion ratio estimated by the MEGA vers. 2.0 (Kumar et al. 2001),
and clades were evaluated by bootstrapping with
115
Lee et al. -- Population Genetics of Giant Wood Spiders
1000 replicates using unweighted characters. The
MP analysis was performed by bootstrapping
(Felsenstein 1985) with 1000 replicates using
unweighted characters, with ACCTRAN, MULPARS, and TBR options, and a heuristic search
with 1 random entry of data by stepwise addition
options with the 50% majority-rule consensus in
PAUP vers. 3.1 (Swofford 1993).
Population genetic analyses
Populations of Nephila pilipes in Taiwan were
comprehensively and systematically sampled to
determine their genetic structures. Specimens collected from 15 populations in Taiwan were grouped
into 6 regions: northern (N, localities 7 and 8),
western (W, localities 9-11), southern (S, localities
12-14), southeastern (SE, localities 15 and 22),
eastern (E, localities 19-21), and northeastern (NE,
localities 16-18). However, since the sample size
from mainland China was small, those spiders
were not included in the population genetic analy1
5
China
4
6
2
3
Taiwan
8
10
7
16 18
17
9
21
11
20
19
13
12
22
23
15
24
sis. Genetic polymorphism and diversity of regions
and populations were quantified by values of haplotype diversity (h) and nucleotide diversity (π)
(Nei 1987). Haplotype diversity was calculated by
the equation, h=n(1-Σf i2) / (n-1) , where fi is the
frequency of the ith haplotype, and n is the sample
size. Nucleotide diversity (π) was estimated by
the equation, π=nΣfifipij2 / (n-1), where fi is the
frequency of the ith haplotype, fj is the frequency of
the jth haplotype, pij is the sequence divergence
between the ith of jth haplotypes, and n is the sample size. The calculation used DNA SP vers. 3.14
(Rozas and Rozas 1999) among geographical
regions and among populations. The Analysis of
Molecular Variance (AMOVA, Excoffier et al. 1992)
program in ARLEQUIN vers. 2.0 (Schneider et al.
2000) was used to deduce the significance of geographical divisions both between western and
eastern groups, among and within populations.
The statistics of molecular variance, ΦCT (between
western (populations: 7-10 and 12-14) and eastern
groups (populations: 15-22), ΦST (among populations within groups), and ΦSC (within populations)
were estimated using ARLEQUIN. In addition,
analysis of the frequency distribution of pairwise
differences in mtDNA sequences, know as mismatch distributions, among Taiwanese N. pilipes
specimens was calculated using ARLEQUIN, vers.
2.0 (Schneider et al. 2000). Slatkin and Hudson
(1991) showed that the mismatch distribution of
stable populations has a smooth unimodal distribution when a population has experienced a recent
demographic expansion. In contrast, the distribution is usually multimodal in samples drawn from
populations at demographic equilibrium.
Nucleotide differentiation between populations
was computed using F-statistics (Wright 1951),
which estimates the proportion of total variance
distributed among populations. FST values were
estimated by the equation,FST =1-(Hw / Hb), where
Hw and Hb are the average numbers of pair-wise
differences (polymorphic DNA sites) within and
between populations (Hudson et al. 1992). Gene
flow was estimated as Nfm, which represents the
effective population size of females times the frac,
tion of migrating individuals. In Wright s FST based
models, N fm is estimated as [(1/F ST )-1]/2 (for
mtDNA).
14
Fig. 1. Sampling localities of Nephila pilipes examined in this
study. Locality codes are the same as those in table 1. The
Central Mountain Range is schematically illustrated along the
north-south longitudinal axis of Taiwan.
RESULTS
Divergence of the mitochondrial cytochrome
oxidase I partial sequence
116
Zoological Studies 43(1): 112-122 (2004)
The partial mtDNA cytochrome oxidase I
gene (617 base pairs and 205 amino acids) of
Nephila pilipes from 24 localities in Taiwan, the
Ryukyu Is. and China were successfully sequenced (Table 1). Among 617 sites, 29 sites were
variable, with transitions found at 25 sites and
transversions at 4 sites (Table 2). Eleven haplotypes were identified from 189 specimens. When
the sequences of haplotypes were compared, 2
non-synonymous substitutions were found: (1)
haplotype CN6 nucleotide site 164 (ATT→GTT),
with the corresponding amino acid site of 55 (Ile→
Val); and (2) haplotypes TW1, TW3, CN1, and
CN3 of nucleotide site 488 (GTT→ATT), with the
corresponding amino acid site of 163 (Val→Ile)
(Table 3).
tances between haplotypes of groups A and B and
C correspond to geographic distances.
Results of population genetic analyses
Haplotype polymorphism occurred only in the
western and northeastern regions in Taiwan (Table
4; Fig. 3). The nucleotide divergence among haplotypes was between 0.2% and 3.5%. Hierarchical
calculations (Table 4) revealed that the haplotype
diversity (h) among geographical regions ranged
from 0 to 0.370. The nucleotide diversity ( π )
among geographical regions ranged from 0% to
0.573%. Hierarchical analysis of genetic differentiation using AMOVA indicated that the proportion
of molecular variance did not significantly differ
between eastern and western groups (ΦCT, p >
0.05). However, molecular variance did exist within 2 regions (western and northeastern regions).
In general, there was little genetic differentiation
among N. pilipes populations in Taiwan (Table 5).
Pairwise differences in COI mtDNA sequences
among Taiwanese N. pilipes specimens were calculated using ARLEQUIN, vers. 2.0. The pairwise
differences ranged from 0 to 21. The raggedness
statistic showed that the raggedness index
(Harpending 1994) was not significantly lower than
the expected value (p = 0.73). The mismatch distribution of N. pilipes populations in Taiwan was
multimodal and poorly fit the corresponding distribution expected from populations undergoing rapid
expansion (Fig. 4). Geographical division assessed by FST with the aid of DNAsp (Table 6) also
indicated that if the localities within a region were
Molecular phylogeny based on the COI partial
sequence
One phylogenetic tree each was generated
from the NJ and MP methods. Tree topology
showed that N. pilipes populations in Taiwan could
be separated into 2 major lineages (groups A and
B) and a minor lineage (group C), with high bootstrap values (99% in the NJ tree and 100% in the
MP tree, Fig. 2). Group A consisted of most specimens collected from 23 localities in Taiwan, the
Ryukyu Is., and mainland China (EA, RK, CN2,
CN4, CN5, and CN6). Group B consisted of specimens from southwestern China (CN1 and CN3)
and northwestern Taiwan (TW1 and TW3). Group
C consisted of 3 specimens from northeastern
Taiwan (TW2). In neither tree did genetic dis-
Table 2. Sequence variation of the mitochondrial cytochrome oxidase I partial sequence in Nephila pilipes.
Dots indicate nucleotides identical to those of haplotype EA
Variable site
1
Haplotype 0
3
1
2
3
4
5
6
7
8
9
10
11
EA
TW1
TW2
TW3
CN1
CN2
CN3
CN4
CN5
CN6
RK
G
.
A
.
.
.
.
.
.
.
.
1
0
6
1
2
1
1
5
4
1
6
4
2
0
8
2
2
3
2
3
5
2
5
0
2
5
6
2
6
5
2
8
3
3
3
1
3
4
3
3
5
8
3
6
7
3
7
0
4
0
6
4
3
6
4
6
6
4
7
2
4
8
8
5
0
2
5
1
4
5
2
3
5
2
7
5
6
5
5
7
4
6
1
0
A
G
.
G
G
.
G
.
.
.
.
T
.
C
.
.
.
.
.
.
.
.
G
A
A
A
A
.
A
.
.
.
.
A
.
.
.
.
.
.
.
.
G
.
A
G
.
G
G
.
G
.
.
.
.
A
.
G
.
.
.
.
.
.
.
.
C
.
A
.
.
.
.
.
.
.
.
A
.
.
.
.
.
.
.
.
.
C
A
.
G
.
.
.
.
.
.
.
.
C
.
T
.
.
.
.
.
.
.
.
A
G
.
G
G
.
G
.
.
.
.
A
.
.
.
.
.
.
T
.
.
.
A
.
.
.
G
.
G
.
.
.
.
C
T
T
T
T
.
T
.
.
.
.
A
.
G
.
.
.
.
.
.
.
.
A
.
G
.
.
.
.
.
.
.
.
G
.
.
A
.
.
.
.
.
.
.
C
.
T
.
T
.
.
.
.
.
.
A
.
G
.
.
.
.
.
.
.
.
A
.
T
.
.
.
.
.
.
.
.
G
A
.
A
A
.
A
.
.
.
.
A
.
.
.
.
G
.
.
.
.
.
G
.
A
.
.
.
.
.
.
.
.
T
.
C
.
.
.
.
.
.
.
.
T
.
.
.
.
.
.
.
C
.
.
T
.
C
.
.
.
.
.
.
.
.
A
.
G
.
.
.
.
.
.
.
.
C
.
T
.
.
.
.
.
.
.
.
117
Lee et al. -- Population Genetics of Giant Wood Spiders
treated as a single population, there was no genetic differentiation among different regions. N fm
ranged from 0 to infinity ( ), and FST ranged from
0 to 0.21.
DISCUSSION
Population genetic structure of Nephila pilipes
in Taiwan
Results of various analyses indicate that populations of N. pilipes in Taiwan exhibit little genetic
(a) NJ-tree
EA (158. China, Taiwan)
RK (6. Ryukyu)
57 CN2 (1. Kwangtung)
CN5 (3. Kwangtung)
90
CN6 (3. Fujian)
0.02
CN4 (1. Yunnan)
99
93
63
0.07
A
TW1 (7. Chilung, Foyan shan)
TW3 (1. Shuanglianpi)
CN1 (5. Kwangtung, Yunnan)
B
CN3 (1. Kwangshi)
C
EA
N. clavata
0.06
TW 1
N. antipodiana
TW 2
0.04
(b) MP-tree
7
TW 3
EA (158. China, Taiwan)
RK (6. Ryukyu)
CN2 (1. Kwangtung)
CN5 (3. Kwangtung)
CN6 (3. Fujian)
CN4 (1. Yunnan)
93
differentiation. The EA haplotype was the only
type in 12 of 15 populations sampled in Taiwan
which represented a fixed haplotype in most areas
of Taiwan. The fixation of a single haplotype
across populations usually does not allow discrimination between historical and current gene flow.
However, populations in the western and northeastern regions exhibited haplotype polymorphism.
The level of estimated gene flow between these
regions and other regions were all considerably
high, ranging from 1.92 to 6.09. Therefore, there
seems to be no isolation between populations on
both sides of Central Mountain CMR. Therefore,
while CMR does seem to be responsible for differentiation patterns of many organisms in Taiwan, it
does not seem to be a major geographic barrier for
N. pilipes.
Why CMR represents a major geographic
barrier to many terrestrial animals in Taiwan but
,
not to N. pilipes may be a result of this organism s
unusual dispersal ability. In terrestrial arthropods,
some groups such as insects, mites, and spiders
have developed distinctive dispersal methods by
aerial travel. For insects (Johnson 1960), aerial
8
16
9
A
100
TW3 (1. Shuanglianpi)
62
CN1 (5. Kwangtung, Yunnan)
21
11
TW1 (7. Chilung, Foyan shan)
96
17
10
B
20
CN3 (1.Kwangshi)
TW2 (3. Shuanglianpi)
19
C
N. clavata
N. antipodiana
13
22
12
Fig. 2. Phylogenetic relationships constructed by the (a)
Neighbor-joining (NJ) and (b) maximum parsimony (MP) methods based on the mitochondrial cytochrome oxidase I partial
sequence of Nephila pilipes. Scale bar for the NJ tree represents branch lengths in terms of nucleotide substitutions per
site. Numbers under branches in the NJ tree represent branch
lengths. Numbers at nodes higher than 50% were derived from
1000 replications for both the NJ and MP trees. Numbers in
parentheses correspond to the sample size. (MP tree: CI =
0.975, RI = 0.958, RC = 0.934).
15
14
Fig. 3. Haplotype frequencies of Nephila pilipes populations
examined in Taiwan.
118
Zoological Studies 43(1): 112-122 (2004)
,
tropical spiders differential adaptability to the climate and vegetation existing during glacial periods. During glacial periods, there were savannas
or pine forests in low-elevation areas of Taiwan
(Tsukada 1966). However, N. pilipes can only
inhabit low-elevation tropical or subtropical broadleaf forests, so the environment in Taiwan during
the peak of glacial periods was not suitable for N.
pilipes. In the Quaternary, at the interface
between glacial and interglacial periods when the
climate became warmer and broadleaf forests
began to form while Taiwan was still partially connected with the Asian mainland, a small number of
N. pilipes might have been able to disperse from
China into Taiwan again. Because N. pilipes exhibits strong dispersal ability, the CMR in Taiwan has
dispersal is primarily limited to adult forms. But in
spiders, except in some rare cases, only juveniles
perform aerial travel. Especially in the moreadvanced Araneomorphae, the spread of individuals can be achieved by ballooning (Decae 1987).
Ballooning behaviors exist among species of 22
families of the Araneomorphae, including the
Tetragnathidae (Dean and Sterling 1985, Decae
1987, Greenstone et al. 1987). Ballooning spiders
were reported to be found on ships 300 km from
the nearest land and from heights of up to approximately 5 km (Gertsch 1979). Therefore, it seems
that few geographical barriers can effectively prevent the dispersal of ballooning spiders.
Grant and Bowen (1998) reviewed several
approaches to the general demographic history of
a population from genetic data and proposed that
a population with low haplotype diversity (h) and
nucleotide diversity (π) may have experienced an
extended or severe demographic bottleneck in
recent times. All populations of N. pilipes in
Taiwan exhibit low h (0.370) and low π (0.362%)
values, which correspond to the low haplotype and
nucleotide diversity situation proposed by Grant
and Bowen (1998). How can this observed low differentiated pattern be interpreted in terms of geographical events which occurred during glacial
periods? Why have the same historical vicariance
events during glacial periods generated considerable differentiation in vertebrate populations, but
have generated no such pattern in N. pilipes? In
,
addition to N. pilipes aerial dispersal ability, this
difference may also result from these tropical/sub-
Frequency
0.9
Freq. Obs.
Freq. Exp.
0.6
0.3
0
1
6
11
16
Pairwise differences
21
Fig. 4. Frequency distributions of pairwise sequences
between individual haplotypes for Nephila pilipes populations in
Taiwan.
Table 3. Pairwise comparison of mitochondrial cytochrome oxidase I partial sequences (617 bases)
between 11 haplotypes. Numbers above the diagonal represent percentage differences corrected by
,
Kimura s two-parameter model (Kimura 1980). Data below the diagonal are the numbers of nucleotide substitutions (transition/ transversion)
EA
EA
CN1
CN2
CN3
CN4
CN5
CN6
TW1
TW2
TW3
RK
N. clavata
N. antipodiana
8/0
1/0
7/0
0/1
1/0
1/0
6/0
16/2
7/0
0/1
36/57
34/54
CN1
CN2
CN3
CN4
CN5
CN6
TW1
TW2
TW3
RK
N. clavata
N. antipodiana
1.3
0.2
1.5
1.1
0.2
1.3
0.2
1.5
0.3
1.3
0.2
1.5
0.3
1.3
0.3
0.2
1.5
0.3
1.3
0.3
0.3
1.0
0.3
1.1
0.2
1.1
1.1
1.1
3.0
3.3
3.2
3.5
3.2
3.2
3.2
3.3
1.1
0.5
1.3
0.3
1.3
1.3
1.3
0.2
3.5
0.2
1.5
0.3
1.3
0.3
0.3
0.3
1.1
3.2
1.3
16.8
17.0
16.8
17.2
16.6
17.0
17.0
17.0
17.7
17.0
16.8
15.8
16.2
15.8
16.4
15.6
16.0
16.0
16.2
15.6
16.4
16.0
10.5
9/0
1/0
8/1
9/0
9/0
2/0
18/2
3/0
8/1
37/57
36/54
8/0
1/1
2/0
2/0
7/0
17/2
8/0
1/1
36/57
34/54
7/1
8/0
8/0
1/0
19/2
2/0
7/1
38/57
37/54
1/1
1/1
6/1
16/3
7/1
0/2
36/56
34/53
2/0
7/0
17/2
8/0
1/1
37/57
35/54
7/0
17/2
8/0
1/1
37/57
35/54
18/2
1/0
6/1
37/57
36/54
19/2
16/3
7/1
40/57 37/57
33/54 37/54
35/58
34/55
39/21
119
Lee et al. -- Population Genetics of Giant Wood Spiders
(Toda et al. 1998). The average annual temperature of central Japan is generally about 14 C, but
that of southern Japan is about 18 C (Shen 1996).
Differences in distribution patterns of these 2
organisms suggest that R. limnocharis might be
more tolerant to low temperatures. Therefore, it is
possible that during peak glacial periods, R. limnocharis in Taiwan was able to maintain a viable
density, but N. pilipes in general could not survive
the severe glacial climatic conditions. As the CMR
gradually arose due to the collision of tectonic
plates, R. limnocharis in eastern Taiwan might
have gradually been isolated from western populations and eventually evolved unique alleles.
However, there might have been no isolated N.
pilipes population in eastern Taiwan during glacial
periods, and they only reentered Taiwan at the end
of each glacial period. On the other hand, CMR
may have been an effective geographic barrier to
R. limnocharis, so the gene flow between populations on the 2 sides of the island was severely limited. However, invading N. pilipes could easily
bypass CMR via ballooning and spread to the
eastern part of the island, thus generating a morehomogenous genetic structure.
not constituted a significant geological barrier for
this species. This may explain why the current N.
pilipes population in Taiwan exhibits such low levels of haplotype diversity and nucleotide diversity.
°
Genetic structure of Nephila pilipes and Rana
limnocharis: a comparison
Similar to the giant wood spider, the Indian
rice frog R. limnocharis is also quite abundant in
low-elevation areas of Taiwan. Toda et al. (1997)
examined the genetic structure of R. limnocharis
populations in Taiwan using allozyme electrophoresis and found an east-west differentiation
pattern across CMR. In another study, Toda et al.
(1998) incorporated specimens from southeastern
mainland China in their analysis and found western Taiwanese populations to exhibit similar genetic structures with the mainland populations. Why
do sympatric Taiwanese N. pilipes and R. limnocharis populations exhibit such striking differences in genetic structures? The variation in population differentiation patterns may have resulted
,
from these 2 organisms differential temperature
tolerance and dispersal abilities. Similar to N.
pilipes, R. limnocharis is also widely distributed in
East Asia (Frost 1985). However, while N. pilipes
can only be found as far north as southern Japan,
R. limnocharis is abundant even in central Japan
The origin of Nephila pilipes in Taiwan:
a preliminary assessment
Table 4. Sample size, haplotype diversity (h), and nucleotide diversity (π) of
Nephila pilipes in 6 regions of Taiwan
Region (locality code)
1. Northern region (N) (7, 8)
2. Western region (W) (9, 10, 11)
3. Southern region (S) (12, 13, 14)
4. Southeastern region (SE) (15, 22)
5. Eastern region (E) (19, 20, 21)
6. Northeastern region (NE) (16, 17)
Sample
size
Haplotype
diversity (h)
Nucleotide
diversity (π)
15
34
17
15
16
35
0.00
0.37
0.00
0.00
0.00
0.23
0.00
0.003
0.00
0.00
0.00
0.006
See text for populations included in each region.
Table 5. Results of the AMOVA test examining the distribution of genetic
variance among 14 populations of Nephila pilipes in Taiwan
Source of variation
Between western and
eastern groups
Among populations
within groups
Within populations
Variance
Percent (%)
p
Φst
0.03
3.19
0.28
0.03 (Φct)
0.01
1.16
0.29
0.04 (Φst)
0.78
95.65
0.18
0.01 (Φsc)
°
120
Zoological Studies 43(1): 112-122 (2004)
The distribution of haplotypes in Taiwan and
China provide some preliminary hypotheses
regarding the origin of N. pilipes in Taiwan. In this
study, a limited number of specimens from China
exhibited high haplotype polymorphism. From 30
individuals collected, 7 haplotypes were found.
However, in 153 individuals collected from Taiwan
only 4 haplotypes were identified. Both NJ and
MP phylogenetic trees demonstrated that TW2 is
most closely related to the outgroup, followed by
group B (haplotypes TW1, TW3, CN1, and CN3)
and group A (haplotypes EA, RK, CN2, CN4, CN5,
and CN6). Since TW2 is distantly related to lineages A and B and is more-closely related to the
outgroups, it is possible that it represents an
ancient haplotype. Since EA is the most-dominant
haplotype in Taiwan and is also the most-abundant
haplotype in southeastern mainland China, it may
represent the most recently evolved haplotype that
entered Taiwan at the end of most-recent glacial
period (about 10 000 to 15 000 yr ago.) If after
intensive sampling we are still unable to find haplotypes TW1, TW2, and TW3 in mainland China,
then it is possible that these haplotypes represent
new mutations from ancestral haplotypes of the
Asian mainland and which are isolated from the
original mainland populations. However, in addition to the fixation of new mutations in isolated
subpopulations, the observed differences between
lineages could also have resulted from ancient
polymorphisms (Hayes and Harrison 1992). If the
TW1, TW2, and TW3 haplotypes are found in
mainland China after intensive sampling, then it is
possible that the current population genetic structure of N. pilipes in Taiwan reflects ancient polymorphisms. Although a comprehensive collection
from mainland China has not yet been conducted,
results from mismatch distribution analysis of N.
pilipes in Taiwan suggest that ancient polymorphisms might be responsible for the observed pattern. When a population has recently gone
Table 6. Pairwise FST and Nfm estimates between
geographical regions of Nephila pilipes based on
sequence variations
1. N
1. N
2. W
3. S
4. SE
5. E
6. NE
2. W
3. S
4. SE
5. E
6. NE
0.21
0.00
0.21
0.00
0.21
0.00
0.00
0.21
0.00
0.00
0.06
0.21
0.06
0.06
0.06
1.92
∞
∞
∞
7.63
1.92
1.92
1.92
6.09
∞
∞
∞
7.63
7.63
7.63
through a rapid expansion, its mismatch distribution tends to be unimodal. In contrast, populations
experiencing long periods of equilibrium usually
exhibit a multimodal distribution (Slatkin and
Hudson 1991). The mismatch distribution of N.
pilipes populations in Taiwan is multimodal (Fig. 4),
suggesting that this population has not experienced a recent rapid expansion but has properties
of a stable population. According to these properties, it is reasonable to infer that N. pilipes entered
Taiwan after the most-recent glacial period and
reoccupied this island from a metapopulation in
southeastern mainland China or the nearby southeastern Asian land mass and thus carried the
properties of this relatively stable metapopulation.
According to this scenario, Taiwanese populations
should be regarded as the same population as in
mainland China instead of an isolated population
which expanded from a few ancestral individuals.
On the other hand, although almost all relevant studies have demonstrated that terrestrial
organisms in Taiwan migrated from mainland
China during glacial periods, it is possible that N.
pilipes in Taiwan could have also migrated from
areas such as the Philippines during interglacial
periods. In addition to dispersing through land
bridges formed during glacial periods, small terrestrial animals such as arthropods may also enter
Taiwan by ocean currents. For example, the northernmost island of the Philippine archipelago, Batan
Is. about 90 km from southern Taiwan, and the
speed of the Kuroshio Current is about 0.9 m/s
(Veron and Minchin 1992). Transportation of
organisms could thus easily occur from the northern Philippines to southern Taiwan in less than 2 d.
From the co-occurrence of many unique plant
species between the northern Philippines and
southern Taiwan, it is possible that a branch of the
Kuroshio Current provides a connection between
these 2 areas. In addition, during summer the
South China Current, driven by the southwestern
monsoon, flows northwards into the Taiwan Strait
through the Penghu Channel (see reviews in Chen
1999). Therefore, both the Kuroshio and South
China Currents may provide opportunities for animals to disperse into Taiwan. Since the tropical
flora of Kenting is quite similar to that of the northern Philippines (Shen 1996), it is quite possible
that the forest-dwelling N. pilipes could have been
transported with plants by ocean currents.
Because specimens from the Philippines were currently not available in this study, the role ocean
currents have played in the origin of N. pilipes in
Taiwan could not be evaluated. In future studies, a
Lee et al. -- Population Genetics of Giant Wood Spiders
systematic sampling of populations in the
Philippines and mainland China will help further
assess the origin of N. pilipes in Taiwan.
Acknowledgments: We thank M. S. Zhu, J.
Chen, X. J. Peng, and W. H. Chou for providing
specimens from mainland China and the Ryukyu
Is. We are grateful to comments provided by C. H.
Kuo, S. H. Lee, C. A. Chen, D. Q. Li, and T.
Oshida. Special thanks are given to all the members of the Behavioral Ecology Laboratory,
Tunghai Univ., Taichung, Taiwan for assistance in
logistics, field collection, molecular techniques,
and genetic data analysis. This study was supported by the National Science Council of the
R.O.C. (NSC-89-2311-B-029-002, NSC-89-2311B-029-008) to I. M. Tso.
REFERENCES
Chen CA. 1999. Analysis of scleractinian distribution in Taiwan
indicating a pattern congruent with sea surface temperatures and currents: examples from Acropora and Faviidae
corals. Zool. Stud. 38: 119-129.
Dean DA, WL Sterling. 1985. Size and phenology of ballooning
spiders in Texas. J. Arachnol. 13: 111-120.
Decae AE. 1987. Dispersal: ballooning and other mechanisms.
In W Nentwig, ed. Ecophysiology of spiders. Berlin:
Springer-Verlag, pp. 348-356.
Excoffier L, PE Smouse, JM Quattro. 1992. Analysis of molecular variance inferred from metric distances among DNA
haplotypes: application to human mitochondrial DNA data.
Genetics 131: 479-491.
Felsenstein J. 1985. Confidence limits on phylogenies: an
approach using the bootstrap. Evolution 39: 783-791.
Folmer O, M Black, W Hoeh, R Lutz, R Vrijenhoek. 1994. DNA
primers for amplification of mitochondrial cytochrome oxidase subunit from diverse metazoan invertebrates. Mol.
Mar. Biol. Biotechnol. 3: 294-299.
Frost DR. 1985. Amphibian species of the world: a taxonomic
and geographical reference. Lawrence, KS: Allen Press.
Gertsch WJ. 1979. American spiders. 2nd ed. New York: Van
Nostrand Reinhold.
Grant WS, BM Bowen. 1998. Shallow population histories in
deep evolutionary lineages of marine fishes: insights from
sardines and anchovies and lessons for conservation. J.
Hered. 89: 415-426.
Greenstone MH, CE Morgan, AL Hultsch, RA Farrow, JE
Dowse. 1987. Ballooning spiders in Missouri, USA and
New South Wales, Australia: family landmass distribution.
J. Arachnol. 15: 163-170.
Hall TA. 1999. BioEdit: a user-friendly biological sequence
alignment editor and analysis program for Windows
95/98/NT. Nucl. Acids Symp. Ser. 41: 95-98.
Harpending RC. 1994. Signature of ancient population growth
in a low-resolution mitochondrial DNA mismatch distribution. Hum. Biol. 66: 591-600.
Hayes JP, RG Harrison. 1992. Variation in mitochondrial DNA
and the biogeography history of woodrats (Neotoma) of
the eastern United States. Syst. Biol. 41: 331-344.
121
Hsu FH, FJ Lin, YS Lin. 2000. Phylogeographic variation in
mitochondrial DNA of Formosan white-bellied rat
Niviventer culturatus. Zool. Stud. 39: 38-46.
Hsu FH, FJ Lin, YS Lin. 2001. Phylogeographic structure of the
Formosan wood mouse, Apodemus semotus Thomas.
Zool. Stud. 40: 91-102.
Hudson RR, M Slatkin, WP Maddison. 1992. Estimation of levels of gene flow from DNA sequence data. Genetics 132:
583-589.
Johnson CG. 1960. A basis for a general system of insect
migration by flight. Nature 186: 348-350.
Kimura M. 1980. A simple method for estimating evolutionary
rates of base substitutions through comparative studies of
nucleotide sequences. J. Mol. Evol. 16: 111-120.
Kumar S, K Tamura, IB Jakobsen, M Nei. 2001. MEGA2:
Molecular Evolutionary Genetics Analysis software,
Bioinformatics. Arizona State University, Tempe, Arizona,
USA.
Lee CL. 1964. The spiders of Taiwan. Taipei: Da-Jian Publishing. (in Chinese)
Lin CC. 1963. Quaternary in Taiwan. Reports on histogeographical studies of Taiwan. Taiwan Wen Hsien 14: 1-92.
(in Chinese)
,
Lin CC. 1966. An outline of Taiwan s Quaternary geohistory
with a special description of the relation between natural
history and cultural history in Taiwan. Bull. Dept.
Archaeol. Anthropol. ROC 28: 7-44. (in Chinese)
Lin CH, CN Lee, CC Yang, BW Lue, SH Chen, K Fang. 1999.
RAPD-estimated genetic relationship of Psechrid spiders
(Araneae: Psechrus) in Taiwan. Biol. Bull. Natl. Taiwan
Normal Univ. 34: 95-104. (in Chinese)
Lin JY, LK Lin. 1983. The preliminary study of animal geography of mammal in Taiwan. Ann. Taiwan Mus. 26: 53-62.
Liu KC. 1995. Phylogeographic relationship of Japalura swin,
honis based on analysis of mtDNA. Master s thesis,
National Sun Yat-Sen Univ., Kaohsiung, Taiwan. (in
Chinese)
Lue KY. 1998. The study of evolution of terrestrial organisms in
Taiwan. Life Sci. News NSC 12: 3-6.
Nei M. 1987. Molecular evolutionary genetics. New York:
Columbia Univ. Press.
Nicholas KB, HB Jr Nicholas, DW II Deerfied. 1997. GeneDoc:
analysis and visualization of genetic variation. EMBNEW.
News 4: 14.
Rozas J, R Rozas. 1999. DNA SP vers. 3: an integrated program for molecular population genetics and molecular
evolution analysis. Bioinformatics 15: 174-175.
Saitou N, M Nei. 1987. The Neighbor-joining method: a new
method reconstructing phylogenetic trees. Mol. Biol.
Evol. 4: 406-425.
Sambrook E, F Fritsch, T Maniatis. 1989. Molecular cloning: a
laboratory manual. 2nd ed. New York: Cold Spring
Harbor Laboratory.
Schneider S, R David, L Excoffier. 2000. ARLEQUIN vers. 2.0:
A software for population genetics data analysis. Univ. of
Geneva, Switzerland: Genetics and Biometry Laboratory.
Shen CF. 1996. The biogeography of Taiwan: 1. Background.
Ann. Taiwan Mus. 39: 387-427.
Shen CF. 1997. The biogeography of Taiwan: 2. Some preliminary thoughts and studies. Ann. Taiwan Mus. 40: 361450.
Slatkin M, RR Hudson. 1991. Pairwise comparisons of mitochondrial DNA sequences in the stable and exponential
growing populations. Genetics 129: 555-562.
,
Swofford DL. 1993. User s manual for PAUP vers. 3.1:
122
Zoological Studies 43(1): 112-122 (2004)
Phylogenetic Analysis Using Parsimony. Champaign, IL:
Illinois Natural History Survey.
Thompson JD, DG Higgins, J Gibson. 1994. Clustal W:
improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, position- specific
gap penalties and weight matrix choice. Nucl. Acids Res.
22: 4673-4680.
Toda M, M Nishida, M Matsui, KY Lue, H Ota. 1998. Genetic
variation in the Indian rice frog Rana limnocharis
(Amphibia: Anura) in Taiwan, as revealed by allozyme
data. Herpetologica 54: 73-82.
Toda M, M Nishida, M Matsui, GF Wu, H Ota. 1997. Isozyme
variation among Asian populations of the Indian rice frog,
Rana limnocharis (Amphibia: Anura). Biochem. Syst.
Ecol. 25: 143-159.
Tsai CY. 1999. Population phylogeny of Gekko hokouensis
based on mitochondrial 12S rRNA and cytochrome b
,
sequences. Master s thesis, National Sun Yat-Sen Univ.,
Kaohsiung, Taiwan. (in Chinese)
Tsukada M. 1966. Late Pleistocene vegetation and climate in
Taiwan. Proc. Natl. Acad. Sci. USA 55: 543-548.
Tzeng CS. 1986. Distribution of the freshwater fishes of
Taiwan. Ann. Taiwan Mus. 39: 127-146.
Veron J, P Minchin. 1992. Correlations between sea surface
temperature, circulation pattern and the distribution of hermatypic corals in Japan. Continent. Shelf Res. 12: 835837.
Wang JP, KC Hsu, TY Chiang. 2000. Mitochondrial DNA phylogeography of Acrossocheilus paradoxus (Cyprinidae) in
Taiwan. Mol. Ecol. 9: 1483-1494.
Wright S. 1951. The genetical structure of populations. Ann.
Eugenet. 15: 323-354.
Yaginuma T. 1986. Spiders of Japan in color. Osaka:
Hoikusha Publishing.
Yang CC. 1999. Molecular phylogenetic determination of
,
Psechrus spiders in Taiwan. Master s thesis, National
Taiwan Normal Univ., Taipei, Taiwan. (in Chinese)
Yang YJ, YS Lin, JL Wu, CF Hui. 1994. Variation in mitochondrial DNA and population structure of the Taipei treefrog
Rhacophorus taipeianus in Taiwan. Mol. Ecol. 3: 219228.
Yeh WS. 1997. Phylogeographic structure of Rhacophorus
,
moltrechti. Master s thesis, National Taiwan Univ.,
Taipei, Taiwan. (in Chinese)